LOL. Earlier, it was suggested that it means 'Half The Photons' - that seems to be Mikael's definition, anyway.
'Half The Photons' is correct IF you concede that enabling HTP raised the effective ISO amplification +1EV from whatever ISO you were at before activating HTP.
so the resultant metering of course causes a -1 EV exposure - and HALF THE PHOTONS!
Sorry, but you seem to be falling into the same trap and suffering from the same misconceptions as Mikael.
It's only half the photons if the aperture is set a stop narrower or the shutter speed is set a stop faster. Are you suggesting that enabling HTP directly affects aperture or shutter speed?
In fact, enabling HTP does nothing to the resultant metering - the camera meters for the ISO you select, it doesn't indicate that it's actually exposing at a 1-stop lower ISO, but it is (and that's why ISO 100 is unavailable with HTP enabled). Try it - set ISO 200 or higher (but not a H expansion, since those are unavailable with HTP), meter a scene, then toggle HTP and see if the metering changes - does it? Even in P Mode with Auto ISO where the camera is selecting all the parameters, enabling HTP does not change the aperture, shutter speed, or ISO selected.
HTP doesn't raise the ISO, unless you have it set lower than ISO 200 before enabling HTP (and if you're at an H expansion ISO, enabling HTP lowers it). I concede that in the specific case of the camera being set to ISO 100 before enabling HTP, the metered exposure will change by one stop, and IF you are in an auto exposure mode or you manually adjust exposure to compensate by one stop, then the number of photons will be halved. Also 'halved' applies only if ISO 100 was set - if it was set to 50, 125, or 160, the change will be more or less than one stop (but not 'halved'). But that's a byproduct of how HTP works, applicable in a limited range of circumstances. To conclude that reducing the photons by half is the mechanism by which HTP works is like concluding that a stopped analog clock is keeping correct time because you just happened to look at it at whatever time at which it was stuck.
As for misunderstanding Mikael, as I indicated earlier and Meh just wrote, that's clearly not the case. He specifically stated, at least five times in at least four different posts, that HTP halves the number of photons and/or the amount of light hitting the sensor, and used that phrasing in describing the mechanism of HTP. That manifestly demonstrates that his understanding of how HTP really works is seriously flawed, and if you are supporting that explanation, you are also failing to understand how HTP works.
Actually, I think all we experienced types know how HTP works as many have amply explained it.
I think this very long argument is in the semantics of describing the comparison of HTP to non-HTP is where I see the confusion.
Now I'm trying to explain what I think Michael was trying to explain regarding HTP and losing half the photons and how this argument got started.
Like Suede, I read enough of this thread to see that all of you experienced tech guys DO understand HTP but I think your methods of explaining it are being misunderstood by some. This is quite possibly a result of different linguistic syntax, typos, or lack of clear explanation by someone who should have been getting more sleep.
Michael, correct me if I am wrong.
I think Michael WAS basing his explanation on the base of 100 ISO, and that enabling HTP now means you no longer have it available so your HTP base ISO is now 200 ISO and, therefore, you'll be using half the photons to create the same tone you would have if you were still using 100 ISO.
I know you all understand this and you as well conceded this in this specific case in your 2nd last paragraph, above.
This is the foundation of how HTP works. Agreed, yes?
Because, regardless of this argument, we all know that one full step up in ISO means we'll now be using half the photons to create a given tone or shade.
I think this got out of context when thinking about the higher ISO settings, with and w-o HTP
E.G. If we take a normal 800 ISO setting and require 100 photons per pixel to generate some shade.
at 1600 ISO we should only require 50 photons to generate the same shade or 200 photons at 400 ISO.
we're still good to this point too, yes?
now we turn on HTP and set to ISO 800
The metering is the same as for the normal ISO 800, hence the exposure is the same and we still use 100 photons to create the same shade.
But behind the scenes, inside the camera, the ISO (sensor gain) is actually ISO 400, and the subsequent processing of this raw file is tweaked so that those 100 photons create the same shade as they did with non-HTP 800 ISO.
The HTP advantage is that since the shot was really done, internal to the camera, at a 400 ISO sensitivity there's 1EV more headroom in the highlights available and the processing allows for a more gradual transfer curve near the top EV levels to prevent or reduce the likelihood of clipping bright shades of gray-white.
The HTP disadvantage is that this comes at the expense of signal to noise ratio which is most evident in shadow areas where fixed read noise becomes a considerable amount relative to the small signal of darker tones/shades.
we're still good ?.. I know what I'm talking about?..
I think Michael merely views an HTP ISO 800 as ISO 400, 1600HTP as 800, etc.
I think his viewpoint is from INSIDE the sensor, and this does not surprise me considering how attentive he is to sensor metrics.
If you view things in this way then you ARE using half the photons compared to using the normal ISO and exposing appropriately.
If you view things in this way then you ARE changing the metering and therefore the exposure and the number of photons/electrons by -1 stop.
When using HTP you ARE getting more shadow noise for a given shade because it's only using half the photons to create compared to using the one step lower non-HTP ISO.
I think perhaps Michael did not succeed in making his point of view clearly to those of us who look at this from a different viewpoint is all.
If I had to explain this in another language i don't think i'd do a very good job of it myself.
I hope I presented a clear explanation of what I think is Michael's viewpoint.
Did I help clear anything up for anybody?