Pages: 1 2 3 [4] 5 6 7   Go Down

Author Topic: Sony A900  (Read 64904 times)

andyptak

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 457
Sony A900
« Reply #60 on: December 10, 2008, 01:49:23 pm »

How about the Sony (non Zeiss) 11-18 mm zoom. Apart from what I guess will be heavy vignetting, is it any good? Thanks.
Logged

michael

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 5084
Sony A900
« Reply #61 on: December 10, 2008, 01:51:18 pm »

Don't know. It doesn't interest me because it's not full frame.

Michael

Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Sony A900
« Reply #62 on: December 10, 2008, 03:37:10 pm »

Hi,

I have that lens in "Konica Minolta" guise. It's pretty bad on my A700. I certainly would not recommend it for the A900.

Erik


Quote from: andyptak
How about the Sony (non Zeiss) 11-18 mm zoom. Apart from what I guess will be heavy vignetting, is it any good? Thanks.
Logged
Erik Kaffehr
 

ziocan

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 426
Sony A900
« Reply #63 on: December 10, 2008, 03:40:47 pm »

Quote from: rljones
I've used the Sigma 12-24 on the A900 and sold it. mushy, low contrast corners. the older Minolta 20/2.8 on the used market a better option on the A900.
I agree it is a good lens indeed.
I have the new version made by Sony. It has one con though, which is a significant vignetting at wide open, which recede pretty much at f4. A part for that is a very good lens with good sharpness over all including the edges on FF and very little CA. Its AF is very precise as well. I think it is better than the Canon 20mm which is not as sharp at the edges and deliver a little more CA.
The vignetting wide open is very strong on the a900 e specially if the scene include white walls and it is bearable on the a700.

Logged

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Sony A900
« Reply #64 on: December 10, 2008, 03:50:10 pm »

Hi,

I have some "first impressions" of the KM 28-75/2.8 on the A900 which I had about one day. Weather here is ugly and I work full time, need to earn some Swedish Kronor to pay for my stuff, so I just did some bookshelf shots. The 28-75 (which is essentially a Tamron) is pretty good. Great surprise was that the Minolta 20/2.8 I had on pasture for three years seem to be a very good performer. I'll post more info once I get around to take pictures for real. I have an SAL 28-70/2.8 ZA on order with promised delivery end of february (ouch!), glad that I have some old stuff. I have also an Sigma 12-24 on order but I may cancel, now that I found that the 20/2.8 is so good.

I would be interested to hear which lenses Michael Reichman bought and the reasons behind his choices.

Quote from: michael
Don't know. It doesn't interest me because it's not full frame.

Michael
Logged
Erik Kaffehr
 

ErikKaffehr

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 11311
    • Echophoto
Sony A900
« Reply #65 on: December 10, 2008, 03:55:00 pm »

I hope that your expedition to Antartica will be a great sucess! Good luck to all participants and crew!

Erik

Quote from: michael
Let's put it this way - I have now purchased two A900 bodies and 5 lenses and intended on using it as my main camera system for the next while.

It's what I'll be taking to Antarctica next month.

Michael
Logged
Erik Kaffehr
 

charleski

  • Newbie
  • *
  • Offline Offline
  • Posts: 34
Sony A900
« Reply #66 on: December 11, 2008, 12:30:57 pm »

Quote from: ejmartin
False.   Dynamic range in stops provides an upper bound to the required RAW bit depth:

http://theory.uchicago.edu/~ejm/pix/20d/te...e/noise-p3.html
The practicalities of analogue to digital conversion are, unfortunately, a little more complex. Although modern sensors have sophisticated normalisation circuitry, they're still subject to fluctuations in the trim of their voltage levels, because of temperature, age and other factors. Since none of the major companies ship their cameras with an engineer inside to check levels for you, this can be an issue. If you have a sensor capable of 11.5bits of DR feeding a 12bit recording format, then a small drift in the DC offset will lead to clipping and loss of DR. The extra headroom afforded by a 14bit format simply allows more leeway for the inevitable fluctuations that will occur in an individual sensor's performance under different conditions. At this level of performance, it's definitely desirable.
Logged

douglasf13

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 547
Sony A900
« Reply #67 on: December 11, 2008, 04:04:55 pm »

Quote from: charleski
The practicalities of analogue to digital conversion are, unfortunately, a little more complex. Although modern sensors have sophisticated normalisation circuitry, they're still subject to fluctuations in the trim of their voltage levels, because of temperature, age and other factors. Since none of the major companies ship their cameras with an engineer inside to check levels for you, this can be an issue. If you have a sensor capable of 11.5bits of DR feeding a 12bit recording format, then a small drift in the DC offset will lead to clipping and loss of DR. The extra headroom afforded by a 14bit format simply allows more leeway for the inevitable fluctuations that will occur in an individual sensor's performance under different conditions. At this level of performance, it's definitely desirable.

  Considering how most evidence of the D300/D3x 14bit mode is pointing towards oversampling of the sensor's 12bit ADCs, I'm not sure how your post applies to those cameras??
Logged

Dan Wells

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1043
Sony A900
« Reply #68 on: December 11, 2008, 07:35:22 pm »

Michael-
    Are you replacing your Phase One gear with the Alpha (is it that good?), or are you using it as your primary small-format system? I'm trying to decide ,as a landscape shooter, between making the jump to MF or buying into one of the super high-res FF systems. I have Canon gear now, which I'm not ergonomically that satisfied with (I have one hand, and I don't especially love the controls - bought into Canon when there was no other choice for image quality in a 35mm body), and I've pretty much decided to switch either to the Alpha, the D3x (I really love Nikon handling), or medium format. If the Alpha is good enough to cause you to put down your Phase system, that's a pretty strong endorsement to the small-format approach in general, and to the Alpha in particular.

                                       -Dan
Logged

Quentin

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 1222
    • Quentin on Facebook
Sony A900
« Reply #69 on: December 11, 2008, 07:47:19 pm »

A good sample of the Sigma 12-24 is worth having.  The design is brilliant - outstanding correction, hardly any CA - but in a poor sample, sharpness is not good.  I don't have one for the A900, but I did use a decent sample for several years on a Kodak 14nx.  

Quentin
Logged
Quentin Bargate, ARPS, Author, Arbitrato

michael

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 5084
Sony A900
« Reply #70 on: December 11, 2008, 08:20:28 pm »

Dan,

By no means am I replacing my Phase back. In fact I'll be upgrading my P45+ to a P65+ when it becomes available in a month or so.

The Sony A900 system is what I'll be using in the DSLR catagory, but for what I do it does not replace the H2 and Phase back.

I should mention for the umpteenth time that what I do in terms of equipment is not at all typical of what others may do. I review gear for my site and magazine articles for a living. This means long term testing. As a consequence when something interests me I will buy it, using it for a time, and then sometimes sell it. The subsequent loss is a cost of doing business. Think Consumer Reports.

Most other sites get a camera sample for a few days or weeks, and then they're gone. Not all, but most. I sometimes do that as well, but not always. Things are never simply black and white.

On the other hand when gear satisfies me, I buy it and then keep it for a long time. This allows me, I believe, a unique perspective when reviewing. I often have many months or years of experience with a product and ten of thousands of frames.

Right now I've had my Hassy system for about 4 years and I'm on my third Phase back. I may switch to a Phase camera in the months ahead, but I'm not sure of that at the moment.

Michael
« Last Edit: December 11, 2008, 08:22:42 pm by michael »
Logged

Kenneth Sky

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 463
    • http://
Sony A900
« Reply #71 on: December 11, 2008, 08:55:29 pm »

Michael
It is my understanding that the reason you have purchased 2 A900 bodies is to take them to the Antarctic. It would be helpful if you could report on how well the "weathered" the in the field use. My use of the A900 with the CZ 24-70, G 70-300 and Sigma 12-24 has been limited to less extreme conditions with no untoward effects. Although the body is well sealed, I still have reservations about the resistance to moisture of the lenses. I know they're not made of sugar but it would be interesting to know how well the system fares in rain or when you wallow around in the snow and ice to get the right  angle.
Ken
Logged

douglasf13

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 547
Sony A900
« Reply #72 on: December 11, 2008, 10:56:48 pm »

Quote from: Dan Wells
Michael-
    Are you replacing your Phase One gear with the Alpha (is it that good?), or are you using it as your primary small-format system? I'm trying to decide ,as a landscape shooter, between making the jump to MF or buying into one of the super high-res FF systems. I have Canon gear now, which I'm not ergonomically that satisfied with (I have one hand, and I don't especially love the controls - bought into Canon when there was no other choice for image quality in a 35mm body), and I've pretty much decided to switch either to the Alpha, the D3x (I really love Nikon handling), or medium format. If the Alpha is good enough to cause you to put down your Phase system, that's a pretty strong endorsement to the small-format approach in general, and to the Alpha in particular.

                                       -Dan

Dan, you should try handling the A900, as everything is adjustable on the camera with only the right hand.
Logged

charleski

  • Newbie
  • *
  • Offline Offline
  • Posts: 34
Sony A900
« Reply #73 on: December 12, 2008, 12:15:02 am »

Quote from: douglasf13
Considering how most evidence of the D300/D3x 14bit mode is pointing towards oversampling of the sensor's 12bit ADCs, I'm not sure how your post applies to those cameras??
I think there's a little confusion here. I'm not talking about oversampling at all (though the benefits of oversampling in both the spatial and intensity domains remains an exciting prospect for future development). I'm talking about simply locating the dynamic range of voltage produced by the sensor within the window of the A/D converter. Ideally (as is the case in professional audio A/D conversion) this is done manually by adjusting the analogue gain on the inputs and ensuring that the DC offset is trimmed. Obviously that isn't an option for photography in the field, so the circuitry needs to rely on 'automatic' normalisation. Such systems are never perfect, though, so it makes sense to build in a leeway for error. The use of a 14bit format to record a 12bit signal affords enough headroom to ensure that the entire 12bit signal actually does get recorded properly even if the baseline voltage has wandered slightly at the time of recording.

The issue here is headroom, pure and simple. If you're trying to fit 11.5bits of analogue dynamic range into 12bits of digital recorded information then the signal needs to be calibrated to fit the A/D converter's input exactly. If you're trying to fit those same 11.5bits into a record that's 14bits deep then it doesn't matter if you're off by a bit (in both senses).

Headroom is cheap, we have 16 and 32GB flash cards, terabyte hard disks and quality brand-name DVDRs for backup that cost a quarter. Increased headroom is a cheap way of making sure that you won't be let down by momentary fluctuations in the state of the analogue data that comes off the sensor.

[EDIT]Perhaps I should clarify: any claims that using a 14bit recording depth leads to oversampling of the data coming off the D3/D700 are simply wrong (the same goes for Canon 14bit formats). The 14bit recording is merely a safeguard to ensure that the full dynamic range of which the sensor is capable is actually recorded. This data can then be renormalised at the time of RAW conversion. I have no doubt that in the majority of cases a 12bit recording format is quite sufficient (since the vast majority of current delivery formats can accommodate no more than 8bits and the majority of scenes fall comfortably within that range). But if you need to dig down into the file to recover as much data as possible through contrast manipulations then a 14bit recording format will make sure that you retain that data over a range of conditions.
« Last Edit: December 12, 2008, 12:41:04 am by charleski »
Logged

Brammers

  • Jr. Member
  • **
  • Offline Offline
  • Posts: 89
Sony A900
« Reply #74 on: December 12, 2008, 10:46:15 am »

Micheal - have you considered ebaying or otherwise sourcing a Minolta 17-35 f3.5 G lens to tide you over til the Zeiss 16-35?  The feedback on FF so far is very good.
Logged

michael

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 5084
Sony A900
« Reply #75 on: December 12, 2008, 11:09:38 am »

No, not really. I don't shoot wider than 24mm all that often. I can wait for the new Zeiss.

Michael

Logged

douglasf13

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 547
Sony A900
« Reply #76 on: December 12, 2008, 11:43:23 am »

Charleski, I'm not saying that 14bit leads to oversampling. I'm saying that there is very, very compelling evidence to suggest that the D300/D3x don't have true 14bit processing like the D3, but, rather, Nikon takes the 12bit data off of the Sony sensor and resamples it to 14bit, thus resulting in the longer processing speed. The slightly less noisy shadow areas of these came in 14bit mode may be attributed to this oversampling, and not attributed to actual 14bit processing. There are tons of threads about this that likely explain it better than I can
Logged

lovell

  • Full Member
  • ***
  • Offline Offline
  • Posts: 131
    • http://
Sony A900
« Reply #77 on: December 12, 2008, 03:18:18 pm »

Quote from: ejmartin
False.   Dynamic range in stops provides an upper bound to the required RAW bit depth:

http://theory.uchicago.edu/~ejm/pix/20d/te...e/noise-p3.html

I gave a better analogy than your meter stick, which takes into account the difference between accuracy and precision, in a discussion here a while back:

http://luminous-landscape.com/forum/index....mp;#entry202059

False.

DR is a function of the sensor and NEVER the A/D converter.  DR is an analog metric, and NEVER a digital metric.  Bit depth has nothing to do with DR.  Bit depth simply defines the sampling of the DR, which is an analog metric.  

The way DR and bit depth work is that bit depth determines how granular the conversion to digital is.  For a given DR called X, X can be digitized to 8, 12, 14, 16 or 32 bits/channel, and deciding how granular the data is converted to has NO impact on DR.

For example of a Canon 10D offers 7 stops of DR, replacing it's 12 bit A/D converter with a 16 bit A/D will not change the DR.  Now if you changed the sensor so that the new sensor provides 9 stops of DR, then doing this would change the DR (of course).
« Last Edit: December 12, 2008, 03:20:10 pm by lovell »
Logged
After composition, everything else is secondary--Alfred Steiglitz, NYC, 1927.

I'm not afraid of death.  I just don't want to be there when it happens--Woody Allen, Annie Hall, '70s

lovell

  • Full Member
  • ***
  • Offline Offline
  • Posts: 131
    • http://
Sony A900
« Reply #78 on: December 12, 2008, 03:22:46 pm »

Quote from: douglasf13
Charleski, I'm not saying that 14bit leads to oversampling. I'm saying that there is very, very compelling evidence to suggest that the D300/D3x don't have true 14bit processing like the D3, but, rather, Nikon takes the 12bit data off of the Sony sensor and resamples it to 14bit, thus resulting in the longer processing speed. The slightly less noisy shadow areas of these came in 14bit mode may be attributed to this oversampling, and not attributed to actual 14bit processing. There are tons of threads about this that likely explain it better than I can

Sensors are not digital devices.  The output of a sensor is analog data.  Not until the A/D processes the signal does the data become digital.

If you took a old D200, which currently uses a 12 A/D, and changed out the A/D with a new one that provides 14 bit depth, the resulting images would in fact be 14 bit images, eventhough the original sensor is used.

Bit depth is not about sensors.  DR is a function of the sensor.  Bit depth is a function of the A/D converter.
« Last Edit: December 12, 2008, 03:24:36 pm by lovell »
Logged
After composition, everything else is secondary--Alfred Steiglitz, NYC, 1927.

I'm not afraid of death.  I just don't want to be there when it happens--Woody Allen, Annie Hall, '70s

BJL

  • Sr. Member
  • ****
  • Offline Offline
  • Posts: 6600
Sony A900
« Reply #79 on: December 12, 2008, 04:49:29 pm »

Quote from: lovell
DR is a function of the sensor and NEVER the A/D converter.  DR is an analog metric, and NEVER a digital metric.  Bit depth has nothing to do with DR.  Bit depth simply defines the sampling of the
Semantics aside, quantization to a given number of bits adds quantization noise to the signal, and if the bit depth is too low, it reduces the DR of the resulting digitized signal. A/D conversion has a quantization noise floor of about 1/2 the value of the least significant bit, and with 12-bit this is at least 1/2^13 of the maximum digital signal level, limiting DR to 2^13 or about 8000:1, no matter how high the DR was in the analog signal entering the ADU.

But with the analog signals of most or all sensors (all except maybe that of the D3 and D700 and some in MF?) apparently limited to DR of  about 4000:1 or less, this 8000:1 DR limit imposed by quantization to 12-bits is probably low enough.

It has been said that ADU's can have noise levels significantly higher than the limit of quantization noise alone, so a particular 12-bit ADU might limit the signal to less than 4000:1. If so, this seems to be a problem with using the wrong 12-bit ADU (one giving less than 12 "significant" bits of output) rather than an inherent disadvantage of 12-bit A/D conversion: the solution is to use an ADU that gives at least 12 valid bits, which should be doable by a good enough 12-bit ADU, though a 14-bit ADU is another option of course.
Logged
Pages: 1 2 3 [4] 5 6 7   Go Up