S1H new price

hojomo said:
Is Panasonic going to announce the S2H in August & is anyone anticipating a (nice condition) used S1H to drop below $3kUS avg before next year?
Click to expand...

I'd say that with the new FW upgrades the S1 offers by far the best bang for the buck for video in the S series. I just picked a used one up for GBP1025 - which was an unusually low price, they're generally GBP1300-1500 body only - and I couldn't be happier!
 
DLD said:
The way I'd look at it is that, switching or not, a decent FF camera for $800 will kill the existing APS-C models until there's something better available. Sony's rumored to have A7IV come out at some point for $2,500 but they might need more than a semi-outdated A7III for $1,500 or so. A7RII is a great buy for stills at $1,200 though.
Click to expand...

Most of it's just academic to me. But I'm very interested in an A7IV. But only if it has 4K60p 10-bit without a crop, plus an adequate readout speed. Otherwise maybe I'll just get a 2nd A7SIII, which might be the way to go anyway. There was an 8 rating used A7SIII for $2750 at b&h recently that sold in a day.

The only thing about the R3 is that if it has an integrated vertical grip, that's a very annoying form factor for video, especially single-handed gimbals
 
A7IV supposedly will not record 8K (not a huge surprise) but will have around 30-32 MPX sensor (a tad on a low side). Its 4K better be real good.

Z9 will have 8K and 16 bit Raw for stills. 45 MPX, 30 fps, ~ $6000-$7,000. And not a low light maven. Its intended target is ... not sure. It depends on the 8K codecs.
 
Fuel for DLD's fire: hyper-anal landscape photographer lloyd chambers says you need 100MP+ in FF35 format to cast away digital artifacts (color aliasing, etc). He doesn't think you need that much resolution for delivery (thus you could shoot with a 100MP sensor and record a 25MP JPEG and reap the benefits), but just to get past the digital nastiness that can occur (when not using an AA filter, which I'm sure he's opposed to for his work): https://diglloyd.com/blog/2021/20210...sorNeeded.html
 
I think the "ugly digital stuff" is from bad processing. Which is why Leica (or ARRI) costs more and delivers more.

Then you get to the film. vs. digital debate, such that has been taking place recently on RedUser. To me, film still looks better (given the proper lighting) but Steve Gleason's point remains - shoot in Raw and then mold the image any way you like.

PS. A7RIV has the double-pixel shift shot that purportedly corrects the digital artifacts without delivering a higher resolution. Which brings me back to the future, i.e., the 1980-s, when the compact disc players fought the oversampling battle -- "we're double, we're quadruple, we're 8x". My first CD player in 1987 was going to be an 8x JVC. As soon as I mentioned it to my buddies in the home audio department, they said, "You stupid idiot. Buy Sherwood. It has a Philips AD converter. It sounds so much better than anything made and sold by Sony". And so I did. But most people couldn't care less.

PPS. Sherwood was the most frequently modded CD player of that era. People replaced the capacitors and the power supply and $250 retail unit became a $1,000 mod piece. But, if you had a nice system, it was well worth it. I only realized how inadequate my 1980 system really was.
 
AD converters are one thing, and I can believe that the quality of the parts and the exactness of the manufacture and quality control are crucial. (Tangentially related - Brawley said that the dynamic range advantage of the Alexa 65 could be chalked up entirely to quality control. They only use the best of the best of their sensors to make that model, and only a fraction of the units they produce pass muster.)

But ARRI also uses OLPFs. And again - the landscape photographer I'm referencing isn't interested in OLPFs. He's interested in detail and accuracy. (Incidentally, I don't think that most of his work is that beautiful. But I read his blog because he flogs his lenses and discusses types of aberrations I've never heard of and interest me academically.)

So, he's talking about getting around digital artifacts without softening the image. And I'm sure that he's not totally correct. Moire is a physical phenomenon that even human vision experiences. If you increase your sensor resolution, it will simply occur at a higher frequency of line pairs. But I suppose that if the frequency is high enough that you never observe it in practice, then you're golden.
 
That's right. With 35mm film, the "digital inaccuracies" would simply be a melt blob of indecipherable images. Higher res, higher ISO, higher frame rate (OK, here, film can be cranked) push the need for higher quality processing a notch higher. And then it becomes a spiral - A9 has more juice than 1D X, then 1D XII has more, then A9II, then 1DX III, then A1. And round and round it goes.

BTW, I was watching some Gold Cup (a soccer tournament held in the US on a biannual basis) highlights last night and whatever camera they had in Houston still produced jagged edges from the 720p era of 2007. And I saw it in some European competition highlights also..

PS. While it may be the case with ALexa65, its sensor is just a side-by-side stack of a three twelve year old ALEF's. By now, its yield ought to be almost perfect and even the rejected batches make a negligible overall effect for a ~ $300,000 unit. Canon has an 8K sensor inside a $4,000 body, Alexa 65 is 6.5K. ARRI makes great cams but it knows how to sell their wares too.
 
Tech Radar writes that R3 will have a 45 MPX sensor ... because the writer thinks he read it somewhere. Most rumors had it at 30. On the other hand, if Canon's going to price it around $5,000, they might as well use the R5 sensor and crop it for a higher frame rate. And make a decent jack-of-all-trades with an oversampled 4K, much like R5 HQ (but without 8K). Canon Rumors is probably correct about 30 though. We should see what's up in September.
 
It's mostly a YouTube game now...the gear acquisition, having the latest and greatest.

People don't need new cameras every year for what they shoot which rarely changes, but constantly buy [and sell] them anyway (myself included).
 
I don't think that's entirely true. Specifically, I think that autofocus systems have been steadily improving, and that's something I hear from photographers as a reason that they transitioned to mirrorless. Personally, that's the main thing that would drive me to replace the A7SIII. The AF system is great, but it could be better. There are definitely situations where it struggles.

But sure, as far as the rest of the camera system is concerned... I've seen blog posts comparing the 5dmk2 with the mk4 or maybe it was the R5. And the imagery totally holds up against the new models.

The same thing with Leica. The photos from their 15-year old digital cameras look great. But the boot-up time is sluggish, the buffer is tiny, burst rate is slow, resolution is low... but now that those factors have risen to the point where most people don't need more, I agree that there is less and less of a reason for anyone to upgrade besides GAS
 
The photo AF is already so good though...

It's like...who really needs even better AF? (Meaning are you and/or is your work that important that you do?)

And in what kind of situations?

Portraits, food, products, and landscapes don't - and that's pretty much half the industry.

Obviously sports and nature could, but you will always miss a shot with so much action constantly moving. Even if your camera doesn't, you might (with bad framing).

It's one thing if you're upgrading from a 5D Mark III to a R3, but from something like a R5? Or a 1DX Mark III? Another $5000+?

It's not necessary but so many do because it's so hard to resist, ha.
 
ahalpert said:
I I've seen blog posts comparing the 5dmk2 with the mk4 or maybe it was the R5. And the imagery totally holds up against the new models.
Click to expand...

Absolutely not. Maybe if you get a comparison under best conditions with enough light and controlled dynamic range the lower numbers don't fall behind far. I had all the models in real life advertising jobs for high class clients and I did review the results very detailed. There was big improvement in each model from noise, high ISO and dynamic range.

And what wasn't directly in the image quality: there was a huge improvement in the AF in each step. But in the end it affected the image as well.
 
NorBro said:
The photo AF is already so good though...

It's like...who really needs even better AF? (Meaning are you and/or is your work that important that you do?)

And in what kind of situations?

Portraits, food, products, and landscapes don't - and that's pretty much half the industry.

Obviously sports and nature could, but you will always miss a shot with so much action constantly moving. Even if your camera doesn't, you might (with bad framing).

It's one thing if you're upgrading from a 5D Mark III to a R3, but from something like a R5? Or a 1DX Mark III? Another $5000+?

It's not necessary but so many do because it's so hard to resist, ha.
Click to expand...

The primary failure points are heavily backlit situations, dark scenes, and scenes with a lot of people moving rapidly. Also, the touch tracking tends to shift to adjacent similar-looking elements. Sometimes, I want it to keep focus on a single word or letter in a sign. When it works, it's amazing. More often, it will shift to a different part of the word or a different word. Things like that could probably be improved with superior AI.​

It would also be nice to have smoother transitions of focus. Perhaps the gyro info from the camera could inform the focus system that the camera is moving, and it could shift focus before it realized the subject is soft. There should also be options to throw the focus all the way out in either direction. Lastly, it would be amazing to tell it to shift before a subject enters the frame, as we can do so easily manually. Or if not that, them a seamless way of focusing manually and automatically in alternation.

For most photographers, maybe they are more satisfied. I imagine that backlit and underexposed scenarios bother them all. And if you look at new features like AF tracking for cars/motorcycles, or animal eye-AF before that, there are clearly some common scenarios of subject selection that haven't been addressed yet.
 
Makes sense but that's a lot of use case nitpicking for something that could probably already have a pretty great hit/miss ratio in its current state.

If anything, I would want to be able to adjust all focus in post, including AI keyframing.

More work, but precise, customizable results.
 
The sonar based external AF system worked differently from the sensor based designs. A subject of one's affection wore a small RF reflector somewhere on his body and that identified him as the center of all the attention. LiDAR works differently too - it scans the frame from the center toward the outer edges and from the nearest subject to the frame to the furthest. IIRC, the sonar based pro systems cost tons and tons of money.
 
NorBro said:
Makes sense but that's a lot of use case nitpicking for something that could probably already have a pretty great hit/miss ratio in its current state.

If anything, I would want to be able to adjust all focus in post, including AI keyframing.

More work, but precise, customizable results.
Click to expand...

I encounter AF failures on every extensive b-roll shoot I do with it. So, you could argue that the system fails on 100% of my b-roll shoots.

And it definitely seems like the type of thing AI could tackle. If the camera gets better at interpreting and identifying data, then it wont get tripped up by backlighting and could better pick out a subject in darkness.

Adjusting focus in post has never been offered on this type of camera. And it would probably cost a lot of time in post-production whilst ceding control from the cameraman. And as I recall, at least the early light field cameras had low light sensitivity. But yes, in theory focusing in post could circumvent all these issues. I'm simply talking about incremental improvements to the current system that are probably within reach (and possibly even withheld by the cartel so they have improvements to offer us later). Certainly, a camera with the exact same specs as the A7SIII but a perfect autofocus system would raise my eyebrows.

Generally, it would be nice if AF could interpret every subject I select as a specific object rather than a color/brightness value. And then it could understand how the look of the object changes as I circle around it. Or identify the letter T that I'm focused on. Etc. And certainly, sonnar/lidar wouldn't help with that. Not by itself anyway
 
Last edited:
Honestly, I don't disagree with your thought-process because I was saying the same exact thing in 2016 and AF has improved so much since then - including many new products with "new" features - but I just think we've reached the point where most people truly don't need any more improvements.

Before 100% V/H coverage and hundreds of points and AI and deep-learning and face and eye-AF, there was so much to improve upon.

Although great AF in pure darkness could be something really interesting (like night vision AI), if anyone requires much more than that and the above then I think it's either their work is really complicated or it's user error.
 
NorBro said:
Before 100% V/H coverage and hundreds of points and AI and deep-learning and face and eye-AF, there was so much to improve upon.

Although great AF in pure darkness could be something really interesting (like night vision AI), if anyone requires much more than that and the above then I think it's either their work is really complicated or it's user error.
Click to expand...

Panny has all those features and their AF sucks, all for lack of PDAF. I don't even understand what deep-learning does in this context. How can the machine learn anything without feedback, without being informed of what it did right and what it did wrong.

At the end of the day, you take the tool into the field and it works or it doesn't. As powerful as it is, it still fails me often. Ideally, it would never fail me.

It's not about AF in pure darkness. AF works substantially worse at ISO 12800 than at ISO 640. It's slower, more hunt-y, and more prone to fix on a bright background rather than the subject. And a good exposure at ISO 12800 1/50 f/2.8 or even 1.4 is not that dark. It's dim but comfortable lighting for human vision, and you encounter it all the time in doc/event situations during night interiors. Even day interiors drive me to ISO 12800 if they're not well-lit, though partially because it's cleaner than ISO 8000. But in that case I end up cutting out light and starving the sensor all the same.

My benchmark for these scenarios is my own vision. I can look at a subject on an actually pure dark dance floor who only has rim lighting, and aim to keep their rim in focus. But the camera can't, and it can't handle more common, less extreme scenarios well, either.
 
If you would like your camera to play God then you may have to wait a bit longer.

When human eyes are blinded by bright light or don’t have enough light, our subject/object identification is affected as well, even while having more dynamic range than cameras.

Currently the odds are stacked against it in your unfavorable conditions and it’s unreasonable to think it will perform perfectly 100% of the time when using it for 8-10-12 hours straight at everything you point it while you’re running around at weddings.

___

And it's been like 8 years since reliable AF has made it into a camera.

You don't think Panasonic Corporation could have better AF if they chose to? Lol

Why it sucks is not up to me to answer, but they could program it to suck if they wanted to.
 
You must log in or register to reply here.
Back
Top