The Ultimate AI Photo Culling Face Off | Aftershoot vs. FilterPixel vs. Narrative Select

Поделиться
HTML-код
  • Опубликовано: 2 янв 2025

Комментарии • 41

  • @jbensonphotography
    @jbensonphotography 8 месяцев назад +12

    Thanks so much for the amazing review! Justin the co-founder of Aftershoot here! Would love to get you on our updated version! Looks like the testing is on an older version of the app! We’re always improving and updating every 4-6 weeks so we have definitely added and improved some things! Let us know if you’d like to give it another run with the updates!!!

    • @BritneyClause
      @BritneyClause  4 месяца назад

      Hey Justin, thank you so much for your comment! Sorry for the delay in reply! I definitely want to test the culling again in the update interface. I've been playing with your AI Editing for a couple months now and feel like I need to revisit the culling too for all the updated goodness!

  • @a-muse6567
    @a-muse6567 7 месяцев назад +4

    My findings were the same, insofar as Aftershoot. I used the software for over a year through multiple updates and over time I found I actually spent more and more time manually culling afterwards because I did not trust the AI picks versus manual picks. One example is that the AI culler seemed to prefer an image where the subject was smiling, and the broader the smile, the higher the rating - even if they blinked. As a senior photographer, many of our favorite images are the serious face. The thing I found most helpful was the closeup view of each face in the preview mode, but I still was going through each image one-by-one, so the "time savings'" was minimal. Especially considering I had to wait for the AI to go through the bunch first (although this can be done while you step AFK). Ultimately I decided I would save the subscription fee and do my own culling. I am faster so perhaps the best thing the tool taught me was speed and efficiency.

    • @BritneyClause
      @BritneyClause  4 месяца назад

      Thank you so much for your comment and your personal experience! That's really interesting that the AI seems to latch on to the smile of a subject. I didn't even think of that being a potential parameter of the system. With that in mind, it really feels like the AI is trained to think "photographer say cheese, which makes smile, pick smile" 🤣.

  • @gibytenshin
    @gibytenshin Месяц назад

    Hello from Tokyo! Thank you very much for the great video! You saved my time in choosing one!

  • @joannawheeler8441
    @joannawheeler8441 8 месяцев назад +7

    30:20 cuts to the chase. narrative and aftershoot tie. altho narrative took 3 secs to cull and aftershoot took 4.5 mins. she also says the ai is the same, same pics selected w all the programs. but she doesnt find the softwares useful compared to manual culling. it took her the same amount of time to cull her own images as the 4 min ones. thx for posting!

    • @BritneyClause
      @BritneyClause  4 месяца назад +2

      Thank you so much for the summarizing! I obviously have some difficulties making things concise 🤣 and your comment inspired me to add a conclusion chapter marker to my future videos!

  • @ceciliamarie2375
    @ceciliamarie2375 4 месяца назад +2

    Great videos! Thanks for the info. I think I'm going to go with narrative select. 3 seconds for 400 images is WILD. Massive game changer. Fingers crossed!

    • @BritneyClause
      @BritneyClause  4 месяца назад

      Right! I was honestly so shocked, I thought I did something wrong at first for it to take that little time, haha! Let me now how it goes for you!
      I'm also intrigued by their AI Editing capabilities now, I haven't tested them out but I'm eyeing them up for the future 😃.
      Thanks for your comment!

  • @healinginfluence
    @healinginfluence Месяц назад

    Wonderful video. Thank you.

  • @ez8308
    @ez8308 3 месяца назад

    I've been using Narrative Select for years. Super easy compared to everything else I've tried. I don't use the auto filters, but I use it for quickly manually culling photos

  • @lisataitphotography9720
    @lisataitphotography9720 6 месяцев назад +1

    Appreciate the deep dive. I've been in the free trial for Filter Pixel (which is now back down to 2 weeks) and there's a couple of things that I've found helpful. I do school photography (among other things) and just uploaded 3909 images - it did it's initial cull in 12m which kind of blew my mind. I also found that the duplicates were about 60-70% accurate - a lot of room for improvement - but this helped as it meant I could just take a quick look at the dupes and then use the compare tool to find the ones with the best focus/expression. There were definitely images in the duplicates that didn't belong and I also found it would sometimes select blurry images - not sure why.
    My major question is how long will it take for the AI to learn my culling style so that it becomes more efficient and I can start to trust it more?? That's why we're all here, right?! Culling is the bane of my existence and I was hoping an AI tool would change my life, but given that I'm hesitant to trust another human to do it, I'm nor sure I should trust the AI yet either. 😬

    • @BritneyClause
      @BritneyClause  4 месяца назад +1

      Thank you so much for your comment and personal experience! This is great information! Almost 4000 photos in 12 minutes is so wild and definitely would free up some of your personal time overall. But I feel the exact same as you in that I found there was a lot of room for improvement in duplicate and blurry image picking. I also don't trust anyone else to cull my images and the AI system that I was using for over a year never really seemed to pick up on my culling style. It's unfortunate because culling is the time suck of all time sucks. I'm hoping to revisit these systems in a handful of months to dig into any major updates and to see if anything has improved! Fingers are crossed.
      The changing of the free trial is also a major bummer. I also noticed Narrative took their 6 projects down to 3, which at least they still are project-based, and one could probably figure out if they liked the system in 3 projects. I'm not a fan of the trials going up and down randomly, though. It feels like users can never be too sure what offering they're going to be able to take advantage of!
      Thanks again for your comment! 😄

  • @pianokeys11
    @pianokeys11 Месяц назад

    For me, it’s definitely a question of whether they actually save me any time? If I have to go through every image anyway, it just adds an extra layer of mental complication to be evaluating their picks and the images vs just the images. Perhaps a first pass that just eliminates the unusable ones? But at the end of the day how much time would that even really save?

  • @RedrumVideoProductns
    @RedrumVideoProductns 9 месяцев назад +1

    I wish you would have looked at how the difference System Resources effected culling in Aftershoot. I'm wondering if that shows any differences in selections. If they're all working off of the same/similar AI models, that one at least seems like it has an edge in how heavily it judges photos, versus the others. I would rather an AI help select a few of the potentially best shots versus assume every shot or even broader from an event shoot that I would want to select. I think for me, that aspect is what's most important in combing through photos. I'll go through the rest manually for the officially released album of photos. If there's an increased accuracy in overall best shots, composition wise, that helps at least get those out more immediately following an event.
    I have massive selection bias anxiety and these assist tools are intriguing to me in relieving that part of my being, as I am slow to cull. I do have my doubts though, as I have used FilterPixel so far and I still have to look through everything manually to make sure it's picking those best shots, but it's at least helped me in the way that it categorizes and puts similar images all next to each other, as sometimes I turn away to grab another little moment for a few shots and then turn back, so the timeline of shots gets screwy. It puts groups of images into context and that's helpful for my process. It's something that I feel like LR could have built in at some point, as they move further into implementing AI as well. That is so great for events where your attention gets pulled all over the place.
    The thing I liked best was the keyword tagging. It could look at an image and give a basic tag for what was going on in the moment, like people hugging and dancing, or portraits and group shots, but this is also something I could see LR implementing at some point. If the others are as accurate in this way as well, which I assume they are, as their AI models seem very similar, I may just have to use that Aftershoot trial for my next few events. FP just felt too barebones for me. It felt more like a shell for more to be built into it, but it still has a skeleton at least, which was great play with. Just left wanting more to justify paying. It's also good to see these products trending down in price as well. Affordability is rough with the types of events I work.
    I'm also interested to see where the AI editing aspects of these software models gets to. I think it's a bit overpriced to be training AI to edit like us, but really just have to go through everything still, especially at first, with a low image database to base everything off of. This last year, I experienced a hard drive crash and I had not backed up the whole previous year's photo edits before it suddenly went. I already culled those albums, but would love to feel like I could get those back to being at least somewhat accurately edited again and have that back catalog able to be pulled from again. The addition of AI tags would help me further categorize so that I can just go back and grab that couple's shot or a group without such an extreme level of combing to go find something.
    Thank you for doing this dive! It really helped me in it breaking down with real life experience. Going to try Aftershoot for the next month of events and hopefully I'll be reporting that I'm saving at least my selection anxiety, if not time!

    • @BritneyClause
      @BritneyClause  4 месяца назад

      Thank you so much for your fantastic comment and personal experience! It really gave me a lot to think about, and I think you're right in that testing the System Resources might impact the culling and I should have tested that! You've given me a great idea for when I revisit these systems and their updates in the coming months!
      I have a lot of picking anxiety as well and when I was manually culling, I would oftentimes go back to my unpicked images and just double-check that there wasn't something better. Even those tiny changes in photos would warrant me adding images to my selections even when I felt I was done choosing. It's a curse for our time management, but lovely for our clients who get lots of options, haha!
      The keyword tagging is fabulous! I found that to be a great feature, that I agree could be implemented in programs like Lightroom pretty easily, especially with the big AI updates coming out recently!
      I know it's been a bit since you left this comment, but I'm curious how your time with Aftershoot was and if you tried them out after their major update. The updates to these systems have me curious about any improvements or differentiations in their AI API, and with your suggestion of testing System Resources, I'm definitely eager to dive back into culling (not something I thought I would ever say
      😆)
      Thank you again for your comment!

  • @XxDeathsEnemyxX
    @XxDeathsEnemyxX 9 месяцев назад +1

    As nice as the thought of having an AI cull large weddings, I think for the time being, I will cull my own. Mayyyyybe the editing will be more useful, but we shall see. Oh, and good luck on your channel, and thank you for putting in the work.

    • @BritneyClause
      @BritneyClause  4 месяца назад

      That's exactly how I still feel 5 months on! I'm going to revisit the culling in a few more months, but I just don't see it quite being the best use of my time or money. AI Editing on the other hand has been a massive game-changer that I do enjoy!
      Thank you for your comment!

  • @icontentacademy
    @icontentacademy 8 месяцев назад +1

    Great review. Thanks for the deep dive and detailed information. What are your thoughts on Imagine-AI? Have you looked at that? Also, I'd love to have you on my podcast, Creative Tech Chats, to chat about AI and photography. I'll send you a DM.

    • @BritneyClause
      @BritneyClause  4 месяца назад

      Hey! Thank you so much for your comment! I know it's been a little bit of time, and I've been in such an AI bubble for the last few months, haha!
      I'll have to check out my Imagen AI video to see if you've commented over there, but yes! I now have tried out their system! And the beta of their culling was, I think, in the beginning stages of doing really cool things, but still ran into the same issues that a lot of the AI culling systems run into. There were bits I liked and bits I didn't like. Since it is in beta I tried to keep my head level on the critiques 😬, but I do look forward to the full release!

  • @BryanMiraflor
    @BryanMiraflor 2 месяца назад

    Tried aftershoot. Idk it was annoying. It was like having to cull the cull. What’s the point 😂
    & Thx for your detailed review!

  • @notmyname8527
    @notmyname8527 4 месяца назад +2

    Super expensive. I pay 10US for Lightroom and Photoshop combined. Yet people complaint about Adobe prices.

    • @BritneyClause
      @BritneyClause  4 месяца назад

      Thank you for your comment! I completely agree! Now that these AI cullers are introducing AI Editing Profiles, the pricing COULD make a little more sense. But still, in a world where all of our programs come with a subscription, it's hard to justify adding something new to the mix! Thanks again! 😃

  • @trancer03
    @trancer03 6 месяцев назад +1

    There is no way ai is going to stick to your style of culling. I am also 100% sure you are going to miss some gems of images if you dont even look at your images yourself. Just use photo mechanic like the pro's do. You still cull in like an hour and it's worth it.

    • @BritneyClause
      @BritneyClause  4 месяца назад

      Thank you for your comment! I totally agree! I had to go back into my Photo Mechanic most of the time to pick the gems I felt I was missing because it was much quicker than going into the culling system after my choices were finalized!
      Depending on the system we learn on, that's where we're going to be quicker and just better at making decisions. I'm well versed in PM, so I found that to be my go-to, and I actually picked faster in PM than my AI culler did. But as new photographers enter the game, they may immediately go to an AI culling system and learn to be quicker in decision-making there. I'd be very intrigued to have a cull speed test race with someone who has only used one of these AI systems to cull to see our differences in cull times 😂!
      Thanks again for your comment! It's given me something interesting to think about!

  • @d1m18
    @d1m18 8 месяцев назад +1

    This is not for all photographs sadly. Yet another comparison that only compares tools for Portrait Photography only. Sad

    • @BritneyClause
      @BritneyClause  4 месяца назад +1

      Hey! Thank you so much for your comment! I don't want to exclude anyone from my testing, so I'm interested in knowing what photo/event types would you like to see me test these systems with? I want to be sure I'm being inclusive to all photo types and since I've had my fair share of dabbling in a lot of session types I can definitely try to see if the type of photo you would like to see tested is something I have access to!
      Thanks again for your feedback, and I hope that in the future I can represent everyone in a more well-rounded way!

    • @d1m18
      @d1m18 4 месяца назад

      @@BritneyClause I enjoy wildlife photography including birds. Would be nice to have something that works on those types of photos
      Cheers!!! 🥰

  • @trancer03
    @trancer03 9 месяцев назад +1

    You should give imagen-ai a try

    • @lisataitphotography9720
      @lisataitphotography9720 6 месяцев назад +1

      Have you used this for culling? I hear it's a good app for colour and editing

    • @trancer03
      @trancer03 6 месяцев назад

      @@lisataitphotography9720 I honestly dont see why people would use AI to do their culling, culling is too much a matter of taste, and with something like photo mechanic its really fast. I can cull thousands of images in like an hour. and while i am culling i can see which images are portfolio worthy etc, i mean i cull at the speed i look at them and i need to look at them at some point anyways so i dont see why you would not even look at them all, you might miss some gems in there.

    • @katehavercroft4940
      @katehavercroft4940 6 месяцев назад +1

      I use imagenAi and i love it for editing. But for culling, I find ultimately it's the same as the point she makes at about minute 32 - you still have to go through everything pretty carefully to see what the AI chose at the final picks. I also find it's picking wayyyy to many keepers than I normally would. So it's nice when it verifies the ones I would have kept but it's just as much work as culling it myself if not a bit harder.

    • @BritneyClause
      @BritneyClause  4 месяца назад

      Hello again! I know this comment is a little older, and I apologize for not responding sooner! I have now tested Imagen AI and created a video on it! I'm not sure if you've commented over there, or not, but I did find it to be an interesting system with a lot of room for improvement. I tested out their culling beta and found it to have the same issues that I had with the non-beta systems; HOWEVER, Imagen AI's culler, when used with an AI Editing Profile does allow you to see your editing style on the images before opening them in Lightroom, which I think is really cool. I'm not entirely sure if it's a massive benefit, but it is neat and sometimes, for me, being neat is a thumbs-up moment 😂. I need to work on my standards, haha!

    • @BritneyClause
      @BritneyClause  4 месяца назад

      @@katehavercroft4940 The beta for culling was really cool! I like how it can show you what the images would look like with you Editing profile over top. Ultimately, though, it has the same issues and definitely has a lot of room for improvement. I'm curious about when the culling is fully released what changes there will be, and I'm excited to try it out when that happens! Thank you for your reply!

  • @OpenSourceCitizen
    @OpenSourceCitizen 8 месяцев назад +1

    AI Culling is crazy. I wonder How many people are actually doing it for hi-end portraits? Seems like several large 4k or 8k monitors/tvs would be quick.

    • @BritneyClause
      @BritneyClause  4 месяца назад

      Ohhh, that's a great question! And something I'm also interested to know as well! I'd love to test out the tethered shooting too, unfortunately, I don't shoot anymore, so I'd have to rely on someone else to do that testing, haha! Thank you for your comment!

  • @OpenSourceCitizen
    @OpenSourceCitizen 8 месяцев назад +2

    I heard like 90 to 100% of AI gui apps are using the same backend. Anybody know the actual name of the backend on these AI Cull tools?

    • @BritneyClause
      @BritneyClause  4 месяца назад

      Hey again! I tried to look into the name of the API that they are all using for their AI culling, but I came up empty-handed in my search. I'll have to try again, since I know someone else is interested in this answer too 😋. AI Editing API is a lot easier to find since I think most of the systems are into the idea of that being something they can sell.

  • @OpenSourceCitizen
    @OpenSourceCitizen 8 месяцев назад

    Sounds like AI Culling is not ready yet, untill they improve the cull-ai and the oversight / reveiw interface.

    • @BritneyClause
      @BritneyClause  4 месяца назад

      Agreed! I've also seen continuous updates from these systems, so I'm definitely going to be revisiting them to see if there is any sort of differentiation now!