MicrostockGroup Sponsors


Author Topic: Anybody using ChatGPT for microstock?  (Read 9626 times)

0 Members and 1 Guest are viewing this topic.

Justanotherphotographer

« Reply #25 on: October 24, 2023, 02:09 »
0
Haha, does that make you  the 'believes 100% everything on t.v. and totally obedient to what the newspapers say' guy?...

LOL, Im just the maybe this stuff doesn't need to be injected into every other thread on a microstock forum guy. Another person has just posted, unprompted and completely without irony, a wall of text about how they are racially oppressed, (as a WASP) then has the gall to post back on topic. A repeated refrain from them. If you dont want threads derailed and to stay on topic why post that nonsense in the first place.

Is the thinking really its fine for one person to post massive rants about how down trodden they are as a white male, or for that matter how Covid is a conspiracy, but not fine for anyone to respond?

By posting that stuff you are either asking for a thread to be derailed or, alternatively, that you should get to post divisive opinions and no one else should get to respond in kind. In Petes case often that you should get to post a massive essay reading like a Tucker Carlson section then  say back on topic like you are the adult in the room.

Pete, you even started a thread elsewhere where this nonsense could be posted, why not keep it in the Off Topic section instead of posting about your aggrievement and repression all over the place. Its like youre obsessed with this stuff?


ribtoks

  • Founder of Xpiks
« Reply #26 on: October 24, 2023, 02:20 »
0
Could you guys just stop the useless unrelated to this topic flood?

Focus on ChatGPT please.

Justanotherphotographer

« Reply #27 on: October 24, 2023, 02:57 »
0
Yeah, all this off topic talk is making it much harder for you to farm us for content for your blog ::)

« Reply #28 on: October 24, 2023, 03:01 »
0
ChatGPT is not available in my country. I use other AIs (copy.ai and Bing) to write titles based on keywords. Also, if I get stuck with keywords, I give my list to AI and it can suggest a bunch of new keywords. Then, of course, I have to go through them and pick just a few that make sense.

ribtoks

  • Founder of Xpiks
« Reply #29 on: October 24, 2023, 03:08 »
0
Yeah, all this off topic talk is making it much harder for you to farm us for content for your blog ::)

Nah, I'm just disappointed to experience Godwin's law (https://en.wikipedia.org/wiki/Godwin%27s_law) again and again on the internet, even in threads that I start!

P.S. You are free to compile responses from this thread to a blogpost of your own, if you think this is how it works and that it's so easy ;)

ribtoks

  • Founder of Xpiks
« Reply #30 on: October 24, 2023, 03:10 »
0
how did this topic turn into conspiracy theories!!!?

It's a freaking Godwin's law https://en.wikipedia.org/wiki/Godwin%27s_law in action. Every. Time.

Justanotherphotographer

« Reply #31 on: October 24, 2023, 06:06 »
0
No one called anyone a N**i though. It would be more accurate to coin a term for an "SJW" or "Woke" or "Neo-Marxist" Law. Far more likely someone's gonna get called that nowadays.

Anyway, I have nothing against Pete being an SJW or Neo-Marxist with politics of aggrievement for his racial group. Or that he is woke to the systemic policies of repression of the white man by the new world order. I make no judgement as to the validity of his aggrievement. I can't speak to the suffering he has suffered under the heel of whoever (though I can't remember anyone bringing up his race on this forum other than him? that could be a lapse in my memory?). But does it have to be forced down everyones throat? Can't we live our lives without these woke SJW people and their agendas?

I also dont care if someone else believes in chem trails, that covid and climate change are a lie. Whatever, its just that theres an off topic section especially for this stuff.

« Reply #32 on: October 24, 2023, 07:27 »
+2
No one called anyone a N**i though. It would be more accurate to coin a term for an "SJW" or "Woke" or "Neo-Marxist" Law. Far more likely someone's gonna get called that nowadays.

Anyway, I have nothing against Pete being an SJW or Neo-Marxist with politics of aggrievement for his racial group. Or that he is woke to the systemic policies of repression of the white man by the new world order. I make no judgement as to the validity of his aggrievement. I can't speak to the suffering he has suffered under the heel of whoever (though I can't remember anyone bringing up his race on this forum other than him? that could be a lapse in my memory?). But does it have to be forced down everyones throat? Can't we live our lives without these woke SJW people and their agendas?

I also dont care if someone else believes in chem trails, that covid and climate change are a lie. Whatever, its just that theres an off topic section especially for this stuff.

Ah, so you are just trolling? Okay, well - if you want to get back on topic, do so. You are the one who derailed it. Was it fun?

Get back on topic then.

Or... are you just demonstrating the usage of ChatGPT for producing a response, and showing how useless it can be/the gibberish it can produce ? :) If so, well xribtoks - there's a demonstration for you!

(Haha, just kidding photographer guy - if it actually was a legit response - who knows - you could have used chatgpt for that response - but, if you were just trolling lol, okay - get back on topic then).
« Last Edit: October 24, 2023, 07:52 by SuperPhoto »

Uncle Pete

  • Great Place by a Great Lake - My Home Port
« Reply #33 on: October 24, 2023, 13:36 »
0
No one called anyone a N**i though. It would be more accurate to coin a term for an "SJW" or "Woke" or "Neo-Marxist" Law. Far more likely someone's gonna get called that nowadays.

Anyway, I have nothing against Pete being an SJW or Neo-Marxist with politics of aggrievement for his racial group. Or that he is woke to the systemic policies of repression of the white man by the new world order. I make no judgement as to the validity of his aggrievement. I can't speak to the suffering he has suffered under the heel of whoever (though I can't remember anyone bringing up his race on this forum other than him? that could be a lapse in my memory?). But does it have to be forced down everyones throat? Can't we live our lives without these woke SJW people and their agendas?

I also dont care if someone else believes in chem trails, that covid and climate change are a lie. Whatever, its just that theres an off topic section especially for this stuff.

Ah, so you are just trolling? Okay, well - if you want to get back on topic, do so. You are the one who derailed it. Was it fun?

Get back on topic then.

Or... are you just demonstrating the usage of ChatGPT for producing a response, and showing how useless it can be/the gibberish it can produce ? :) If so, well xribtoks - there's a demonstration for you!

(Haha, just kidding photographer guy - if it actually was a legit response - who knows - you could have used chatgpt for that response - but, if you were just trolling lol, okay - get back on topic then).

Yes, I guess sometimes a bit of humor or sarcasm doesn't come through on a forum. My post was in support of your experience with Chat GPT have a leaning towards revisionist history, meaning, everything now is universal and multicultural and the old days were all controlled by horrible, nasty, mean, invading, racists. My favorite part was "Cultural Appropriation: The appropriation of cultural elements from minority groups by those in power can lead to the erasure of cultural identity and perpetuate harmful stereotypes."

And no I didn't really bring up my heritage? I'm an American. I was born here and so were my parents, But if anyone wants to actually understand, My Grandfather came over on a boat from Italy to escape persecution and poverty, to a better place. He wasn't a mobster, or associated with the Black Hand. And back then, Italians were persecuted in the US, got the low pay and crappy jobs.

"Anti-Italianism arose among some Americans as an effect of the large-scale immigration of Italians to the United States during the late nineteenth and early twentieth centuries. The majority of Italian immigrants to the United States arrived in waves in the early twentieth century, many of them from agrarian backgrounds. Nearly all the Italian immigrants were Roman Catholic, as opposed to the nation's Protestant majority. Because the immigrants often lacked formal education and competed with earlier immigrants for lower-paying jobs and housing, significant hostility developed toward them."

Yeah, get over it. They adapted to the American life and culture and assimilated into society. Oh yeah, they all lived in an Italian neighborhood in Chicago and still spoke Italian at home.  I have cousins that still do. It's a choice.

The other Grandfather escaped from Transylvania, under similar circumstances. Religious oppression, poverty and social stigma, as the Transylvanian Saxons were the lowest of the low.

So next time you make a joke about some gangster in the Mafia, or Italians that are all mobbed up or Dracula being a vampire from Transylvania, I want you to be politically correct and stop. No harmful stereotypes.  You wouldn't want to culturally generalize, stigmatize or appropriate either of those. If you continue my Cousin Mario will come with a baseball bat, and give you a lecture on how to be polite around WOPs and guinea hens.  ;D Or the other kind of bat, will bite you on the neck, and you'll be forever sleeping in a coffin and only able to come out at night. (I never did work out how that garlic thing could be overcome, as just about everything I cook has garlic, oregano and basil in it. But that's supposed to keep away vampires?  :o )

Back on topic, it would be nice if AI could look at an image and say, here are keywords, for us. On the other hand, isn't that what keyword suggestions already do on Adobe, SS and DT, for example?

Midjourney has a nice option, to /describe an image for reprompting. You have to upload an image and you can obtain 4 different descriptions, descriptions that often contain useful specific terms and look good for chatGPT when you ask for keywording.
Unfortunately, the time to:
1- produce a good image
2- upload to midjourney and ask for descriptions
3- check descriptions and use them to obtain good keyword list from chatGPT
4- check the keywording ChatGPT job to find out how accurate is it, what to delete, and the right keywords order in alphabetical order

... is by far longer than simply do the usual semi manual keywording

Is the free Midjourney able to do this?

And just to be clear. Like others, I have standard sets of keywords, I attach them to images from groups, the easiest is my Racing keywords, which get dropped in. But I wouldn't be against at least looking to see what AI has to say. Maybe I could get some more. And last I checked, I removed some that were kind of distant and I had never seen anyone, anywhere, except us stock folks, use as a search or keyword. Why load up images with fluff that serves no purpose?

Does anyone here call an auto race a "motorshow" I'm cutting that one.

Maybe I could try to work around the obvious and use Chat Gpt and ask, "Please name some words that would be associated with motorsports and motor racing"

Glad you brought this up. Something for a snow day.

Uncle Pete

  • Great Place by a Great Lake - My Home Port
« Reply #34 on: October 24, 2023, 13:45 »
0
No one called anyone a N**i though.

That's true, good point.

« Reply #35 on: October 26, 2023, 02:05 »
0
What query did you use to get "good" keywords, and how many were you able to get at a time?

I've found generally speaking the results have been inconsistent, and limited. How were you able to get good results?

It's easy to ask all community than actually use a tool called search for it first isn't it?

Have you tried using your own tip? Did you find in search any results relevant to microstocks?

I'm asking about a personal experience, relevant to microstocks. Not "how to use ChatGPT for photography".

Probably i was not clear since i was trying to make a joke and respond to you at same time what you are asking. But I can guarantee that the joke is not on you but on the situation: it involves AI and the type of responses that Chatgpt sometimes makes if you ask him to play a role.

This time I'm going to try to keep more unboxed: Yes. ChatGPT AI actually creates good keywords. My advice/tip is use it!

However i did try to ask Chatgpt, without role playing,  your question.
Here is the result:

"As of my last knowledge update in September 2021, ChatGPT, like other AI models, was not typically used in the context of microstock photography or similar industries. Microstock refers to the sale of stock photos, illustrations, and other digital media through online platforms, often for a lower cost than traditional stock photography.

However, AI and machine learning technologies have been applied in various ways in the field of photography. This might include image recognition, automated tagging, and content recommendation systems. These technologies can help streamline the process of managing and searching for stock images, but it's typically not the AI model like ChatGPT that is directly involved in these tasks.

If there have been any developments or specific applications of ChatGPT or similar AI models in the microstock industry since my last update, I would not have that information. I recommend checking with the latest industry news or consulting with professionals in the field for the most up-to-date information on AI usage in microstock and related sectors."

I'm sorry to hear that you're not having a good experience with chat gpt but for me its been great. I have been using for almost a year now and I've been getting good results.  Sometimes I change 4 or 5 out of 40. 

It's no big secret sauce but to keep it simple I made a template where I give the instructions. it improves my title and gives me 40 keywords, not just descriptive ones, but conceptual and emotional ones where the clip can also be used. Therefore, not only can be used for documentary but also for storytelling and other point of views. In the instructions I set to do it for 50-100 or different clips at once.

In just over ten seconds I have all the work done by chatgpt where I only analyze and rectify those that I think are exaggerated for the image/video. In less than 15 minutes I made close to 250 different clips where I practically only lost time copy pasting from one side (chat gpt) to the other (website). Probably 15 minutes is overkill... maybe 5-7 minutes will be more realistic.

With the old methods of self-generating in microstock website or using a keywording tools, i will waste more time. much more time! I really recommend it especially for those which English is not the native language. They can write in their own language and have the work in perfect English with good and refined keywords.

Anyway if chatgpt doesn't work for you, then use other "guitar" that works best. ;)
« Last Edit: October 26, 2023, 02:10 by Evaristo tenscadisto »

« Reply #36 on: October 26, 2023, 02:58 »
0
It's no big secret sauce but to keep it simple I made a template
This is interesting!
May I ask you more details about this? How a template can be adapted to many different images (or clips)?
Is it necessary to have the pay version or it can also be done with the free chatGPT?
Thanks

« Reply #37 on: October 26, 2023, 04:51 »
+6
It's no big secret sauce but to keep it simple I made a template
This is interesting!
May I ask you more details about this? How a template can be adapted to many different images (or clips)?
Is it necessary to have the pay version or it can also be done with the free chatGPT?
Thanks

I use the free version. you don't need to pay.
The template is not about code or anything programmed. 

Although these days requesting tasks on chatgpt is a kind of programming the template that i created is just a set of tasks followed by considerations that I ask  chat GPT to take into account before performing the work. Then I add to that the list of 50-100 clips with the names I rendered. Clip names are simply basic descriptions of what the clip contains.

If you are going to create one keep in mind simple tasks with direct instructions. Don't use to much "Considerations" to perform the work. One for technical aspects and one for behavior/subjective aspects is enough. If you put more you start making entropy for chatgpt and you start getting redundancy in the feedback of your keywords. Less is more accurate if well defined at the begin.

In short the template structure is:
1 - Purpose of task
2 - Simple tasks with direct instructions (i only use 5 tasks)
3 - Considerations/work performance (How the work will be done by chatgpt in terms of Form and Content)
4 - List (list of names described. i use 50 or 100)

Hope it helps you.  ;)
 


« Reply #38 on: October 26, 2023, 09:25 »
+2
No one called anyone a N**i though. It would be more accurate to coin a term for an "SJW" or "Woke" or "Neo-Marxist" Law. Far more likely someone's gonna get called that nowadays.

Anyway, I have nothing against Pete being an SJW or Neo-Marxist with politics of aggrievement for his racial group. Or that he is woke to the systemic policies of repression of the white man by the new world order. I make no judgement as to the validity of his aggrievement. I can't speak to the suffering he has suffered under the heel of whoever (though I can't remember anyone bringing up his race on this forum other than him? that could be a lapse in my memory?). But does it have to be forced down everyones throat? Can't we live our lives without these woke SJW people and their agendas?

I also dont care if someone else believes in chem trails, that covid and climate change are a lie. Whatever, its just that theres an off topic section especially for this stuff.

Self appointed moderator control freak has left the forum.

« Reply #39 on: October 26, 2023, 09:41 »
0
Hope it helps you.  ;)

Thank you very much Evaristo for taking time to explain, really interesting! :)

Uncle Pete

  • Great Place by a Great Lake - My Home Port
« Reply #40 on: October 26, 2023, 12:16 »
0
Hope it helps you.  ;)

Thank you very much Evaristo for taking time to explain, really interesting! :)

Ditto to that.

« Reply #41 on: January 30, 2024, 08:59 »
0
I actually use ChatGPT currently to create prompts.
I usually get inspired by an interesting photo/image, write a brief description or some keywords of the content/topic and then rephrase it via ChatGPT.
This saves a lot of time and increases productivity.
Sometimes I even create new and better image compositions.

My perception is also that the more extensive the prompts, the more detailed images can be created with a higher prompt guidance scale in Stable Diffusion.
It's crap to use keywords like "higly detailed", "photorealistic", blablabla, etc. in prompts.

As Evaristo tenscadisto has already written, it makes sense to create templates with short, specific instructions.
Sometimes, for example, I only paste in 3 to 5 keywords and create a detailed prompt because ChatGPT knows from previous example descriptions and specifications how I would write the prompts myself.
As a power user, you could certainly try to automate the whole thing.


« Reply #42 on: January 30, 2024, 09:57 »
0
I actually use ChatGPT currently to create prompts.
I usually get inspired by an interesting photo/image, write a brief description or some keywords of the content/topic and then rephrase it via ChatGPT.
This saves a lot of time and increases productivity.
Sometimes I even create new and better image compositions.

My perception is also that the more extensive the prompts, the more detailed images can be created with a higher prompt guidance scale in Stable Diffusion.
It's crap to use keywords like "higly detailed", "photorealistic", blablabla, etc. in prompts.

As Evaristo tenscadisto has already written, it makes sense to create templates with short, specific instructions.
Sometimes, for example, I only paste in 3 to 5 keywords and create a detailed prompt because ChatGPT knows from previous example descriptions and specifications how I would write the prompts myself.
As a power user, you could certainly try to automate the whole thing.

Interesting... what would be an example of a prompt you use? Playing around with it myself - I found ChatGPT starts to make very similar/redudant stuff very quickly, which almost removes any kind of utility it would have, because have to constantly be checking it to make sure it did things correctly...

« Reply #43 on: January 30, 2024, 10:54 »
0
For example, I am currently trying to generate authentic culture-specific portraits.
The repetitive is sometimes even desired.

I would say this is still the lowest beginner level and can be extended.

I have given ChatGPT some examples of how the prompt should be constructed with a description of the person, culture and landscape in the background.

These are two following generated prompts and generated images in Stable Diffusion.
(raw images are from the first run and must either be adjusted in Stable Difussion by variations or manually retouched after upscaling).

Keywords:
african female, authentic dressed, backdrop sunset savannah

Prompt:
Detailed close up of an African woman adorned with a traditional headwrap, symbolizing humble beauty against the backdrop of an African savanna with a breathtaking sunset. The scene captures the essence of cultural richness and natural splendor, celebrating the woman's grace and connection to her heritage.

Image:
https://ibb.co/brF4dxz

Keywords:
spanish female, authentic dressed, dancing flamenco, backdrop madrid sunset

Prompt:
Detailed Close Up of an elegant, traditionally dressed Spanish woman passionately dancing Flamenco in front of madrid city against the backdrop of a breathtaking sunset. The scene captures the cultural richness and emotive beauty of Flamenco, blending the artistry of dance with the captivating hues of the evening sky.

Image:
https://ibb.co/ZYf4dhW

I currently need about 1 day to generate a certain set-up (approx. 30 to 40 images), sort them out manually and retouch them. That leaves me with about 10 to 15.
I try to generate daily up to 5 max 8 set-ups, so about approx. 50 to max. 120 final images a day. In the meantime I prepare the following prompts.
Currently it is still a very manual and multi tasking task.

But I currently earn a maximum of 10 USD / month with it.
I still have almost 800 images in review on Adobe ...
« Last Edit: January 30, 2024, 11:38 by Andrej.S. »

« Reply #44 on: January 30, 2024, 17:24 »
+1
Hello,

GPT-4 allows you to create a custom version of chatGPT for a specific purpose. I created a custom chatGPT called "Stock Footage Keyword Export". In this chatGPT, I uploaded all the guidelines from various agencies in pdf format, which relate to best practices in keywording, writing titles and descriptions, etc.

I instructed the chatGPT to provide me with 50 keywords and a title for every photo I upload. When I upload a photo from a screen grab, the chatGPT gives me a basic title with many relevant and less relevant keywords. I choose the most relevant keywords and still use other tools such as Wikipedia or SS Keyword tool to find more keywords.

I've attached a screenshot of an example for your reference.

To customize your GPT, I recommend using this comprehensive metadata guide from Adobe Stock https://adobestock.adobe.com/Metadata-Field-Guide.html

Cheers,

Daniel
« Last Edit: January 30, 2024, 18:46 by danielstassen »

Uncle Pete

  • Great Place by a Great Lake - My Home Port
« Reply #45 on: January 31, 2024, 15:18 »
0
Does AI use stop words or ignore them? Does punctuation have any value? Why do I see prompts that have things like f11:10 on the end? What does that mean?

I'd like to know what actually matters, instead of just throwing out words and having some random seed start to make a image that's using some of that and ignoring other parts.

Anyone have a guide?

I've tried a number of times, just for an example, to get SDXL to make "A cow jumping over the Moon" Never one, it puts in a cow, on the moon, over a moon background, all kinds of things, but not a cow, above the moon or jumping over it. DALL-E does better but low quality output.

« Reply #46 on: January 31, 2024, 16:30 »
0
Hello,

GPT-4 allows you to create a custom version of chatGPT for a specific purpose. I created a custom chatGPT called "Stock Footage Keyword Export". In this chatGPT, I uploaded all the guidelines from various agencies in pdf format, which relate to best practices in keywording, writing titles and descriptions, etc.

I instructed the chatGPT to provide me with 50 keywords and a title for every photo I upload. When I upload a photo from a screen grab, the chatGPT gives me a basic title with many relevant and less relevant keywords. I choose the most relevant keywords and still use other tools such as Wikipedia or SS Keyword tool to find more keywords.

I've attached a screenshot of an example for your reference.

To customize your GPT, I recommend using this comprehensive metadata guide from Adobe Stock https://adobestock.adobe.com/Metadata-Field-Guide.html

Cheers,

Daniel

Thanks for the idea. Have seen today some user on reddit, who linked to some self built key word tools with ChatGPT4.
I am currently considering other areas of possible applications and may take a subscription in future.

Does AI use stop words or ignore them? Does punctuation have any value? Why do I see prompts that have things like f11:10 on the end? What does that mean?

I'd like to know what actually matters, instead of just throwing out words and having some random seed start to make a image that's using some of that and ignoring other parts.

Anyone have a guide?

I've tried a number of times, just for an example, to get SDXL to make "A cow jumping over the Moon" Never one, it puts in a cow, on the moon, over a moon background, all kinds of things, but not a cow, above the moon or jumping over it. DALL-E does better but low quality output.

There is basically a prompt syntax for stable diffusion.

- Type of shot (portrait, full body, from behind, macro, etc.)
- Description of the subject
- Possibly control of the depth of field by means of aperture specification or e.g. "bokeh", "shallow depth of field". Mostly with f1.8 or f2.0. Focal length of objects can also be specified, but in my opinion it has no great influence, rather directly control the portrait type such as Extreme Close Up, Close Up, Medium Shot, Full Shot, etc.
- Additional information e.g. on foreground, background, colors, (rarely also composition, image harmony)
- For portraits, additional information on the facial expressions / gestures of the person (smiling, laughing, sad, angry, etc.)

Weighting works via bracketing.
Square brackets [] reduce the weighting, round brackets () increase the weighting.

Commas and periods play a role in separating information and content. Commas are light and periods are strong separators. However, this is actually very similar to normal punctuation when writing texts.
The same applies to the negative prompts.

With LORAS (fine adjustment models), additional weighting can be applied, e.g. with<lora:sdxl_photorealistic_slider_v1-0:1> the number after :

In addition, the images can be further fine-tuned via the number of steps (iterations) and the guidance scale.
With higher settings, the images generally become more detailed, but the prompts must be formulated very precisely.

Basically, I don't think much of controlling the type of image in the standard model by specifying "photorealistic", "artistic", "surreal", etc.
Instead, you should use this in combination with specially optimized models (see https://civitai.com/models).

I believe that open source Stable Diffusion can beat Midjourney in the long term because of the huge developer community and extreme flexibility.
But it's currently really pain in the ass to find right settings, which deliver good results.
« Last Edit: January 31, 2024, 16:35 by Andrej.S. »

Uncle Pete

  • Great Place by a Great Lake - My Home Port
« Reply #47 on: February 01, 2024, 15:14 »
+1

There is basically a prompt syntax for stable diffusion.

- Type of shot (portrait, full body, from behind, macro, etc.)
- Description of the subject
- Possibly control of the depth of field by means of aperture specification or e.g. "bokeh", "shallow depth of field". Mostly with f1.8 or f2.0. Focal length of objects can also be specified, but in my opinion it has no great influence, rather directly control the portrait type such as Extreme Close Up, Close Up, Medium Shot, Full Shot, etc.
- Additional information e.g. on foreground, background, colors, (rarely also composition, image harmony)
- For portraits, additional information on the facial expressions / gestures of the person (smiling, laughing, sad, angry, etc.)

Weighting works via bracketing.
Square brackets [] reduce the weighting, round brackets () increase the weighting.

Commas and periods play a role in separating information and content. Commas are light and periods are strong separators. However, this is actually very similar to normal punctuation when writing texts.
The same applies to the negative prompts.

With LORAS (fine adjustment models), additional weighting can be applied, e.g. with<lora:sdxl_photorealistic_slider_v1-0:1> the number after :

In addition, the images can be further fine-tuned via the number of steps (iterations) and the guidance scale.
With higher settings, the images generally become more detailed, but the prompts must be formulated very precisely.

Basically, I don't think much of controlling the type of image in the standard model by specifying "photorealistic", "artistic", "surreal", etc.
Instead, you should use this in combination with specially optimized models (see https://civitai.com/models).

I believe that open source Stable Diffusion can beat Midjourney in the long term because of the huge developer community and extreme flexibility.
But it's currently really pain in the ass to find right settings, which deliver good results.

I'm going to print that! Your answer led me to search for the correct question, and I found an additional FAQ. Thank You

Syntax for Stable Diffusion  https://www.generativelabs.co/post/prompt-syntax-for-stable-diffusion-faq

I see that the parts I was looking at are not f stops, but numerical weights. Also there's a 75 token limit. So some of the terribly long prompts I've seen could be truncated? Interesting.

I'm using DALL-E and SDXL 1.0
« Last Edit: February 01, 2024, 15:19 by Uncle Pete »

« Reply #48 on: February 01, 2024, 18:45 »
+1
@Uncle Pete
Interesting, I didn't know about the token length restriction in Stable Diffusion. Thanks for the info!
I'll try it out with longer, detailed prompts.

Which online provider do you use?
Or do you generate locally on your own machine?

I currently use Mage.Space. Can't complain. For 15 USD / month unlimited generation with private mode and many pre-installed models / loras.

The cool thing is that in combination with precise prompts and special models, you can maintain character continuity quite well without too much deviation.
So it has potential for image series. I haven't seen sofar Midjourney users using this, but maybe nobody thinks about it.

I'm currently trying to generate colorful portraits with African-American models:
https://ibb.co/PhRSgNN
https://ibb.co/PTkRFZZ

And also the current body positivity, female sexual wellness trend:

https://ibb.co/f1sWkQR
https://ibb.co/1J0f0mM

Oh and yoga in a more "raw" style:
https://ibb.co/nLPSjnw

yes, the hands and minor artifacts are still really annoying at the moment ...

But as I said, I see much more potential in Stable Diffusion than Midjourney to stand out through individualization from the huge same-looking crowd. The buyers will be soon so pissed off because of millions of same looking stuff.

« Reply #49 on: February 01, 2024, 22:21 »
0
I had a look at the mage page.

In the terms of service it says that everything created is public domain.

Not sure if that is ideal.


 

Related Topics

  Subject / Started by Replies Last post
30 Replies
15800 Views
Last post June 09, 2013, 17:19
by cascoly
72 Replies
60667 Views
Last post July 08, 2011, 15:22
by cathyslife
36 Replies
31103 Views
Last post January 10, 2013, 06:03
by Anyka
6 Replies
5329 Views
Last post September 10, 2014, 04:45
by 3Stock
15 Replies
4006 Views
Last post December 13, 2022, 12:29
by Unamas

Sponsors

Mega Bundle of 5,900+ Professional Lightroom Presets

Microstock Poll Results

Sponsors